One-Bottle Preamp

[Kingston]

[EDIT]

I was talking out of my ass.

Here's some new findings,

I measured the output with a spectrum analyser. There was nothing on the input, except a microphone that wasn't on. Otherwise it hums a lot more, is that normal with nothing on the input?.

Output goes straight to RME sound card XLR input (+4db). Preamp had gain set in exact middle position on both channels, no attenuator or pad was used.

Channel1 is 12AV7 and channel2 5965A. Otherwise they are identical.

here is the channel1 spectrum,

http://www.michaelkingston.fi/files/chan1.png

here's channel2.

http://www.michaelkingston.fi/files/chan2.png

I seem to be picking a whole lot of hum. I swapped the tubes around, and the 5965A always puts out that amount of hum.

Oh dear. I checked the connections and they seem alright, complies with all the schematics, so I wonder where that hum is coming from, or what causes it? And why do the tubes behave *that* different? Should there be filtered and regulated DC on the heaters instead of the raw 6.3AC from the PSU transformer?

 

[khstudio]

Are you saying the 5965 has more hum than the 12av7. Have you tried a different 5965. I'm about to start soldering & would like to know. Try swapping channels & see.

Please let us know.

What brand is the 5965?

 

[Kingston]

I *did* swap the channels (like I said), the hum follows. Swap my diagrams around for the same result.

The 5965A is by General Electric, 12AV7 is RCA.

I'll do pictures later to show if there's something particularly wrong or unorthodox going on that might influence these things.

But by all means use the 5965, you might have better luck with it.

 

[NewYorkDave]

You had an open mic cable connected to the input? That'll definitely cause some hum.

The right way to measure the hum/noise level is with the input terminated by a 150 or 200-ohm resistor and NO mic cable connected. Strap the resistor between pins 2 and 3 on the input XLR connector.

Is your heater supply floating--as in, it has no connection to ground or to any DC reference in the circuit? That'll definitely cause hum. Connect a 100-ohm, 1/2-watt resistor (preferably 1%) between each leg of the heater supply and ground. If you don't have the resistors handy, try connecting ONE leg of the heater supply directly to ground. It won't be as quiet as balancing the heaters (as above), but it should reduce hum level compared to floating.

Assuming the layout was done well, there should not be excessive hum even with an AC heater supply. Did you twist the heater supply leads and route them away from signal leads?

The 5965 was developed for computer use but has published characteristics similar to type 12AV7. And although I haven't used it myself, I know that many others have applied the 5965 successfully in audio circuits. Since your problem follows the tube, it stands to reason that there's probably something wrong with that tube--have you tried a different 5965? Is yours used or new-old-stock? If it spent its working life inside a computer, it would have spent long periods of time in cutoff, which will eventually cause a tube to develop noise problems.

Also--and this is just conjecture on my part--5965s may not have been specified as tightly for noise as a receiving type like 12AV7, so it may take more sorting to find a quiet one. At any rate, I don't think we can draw any conclusions based on a lone tube without trying other examples of the same type number.

 

[Kingston]

[quote author="NewYorkDave"]You had an open mic cable connected to the input? That'll definitely cause some hum.[/quote]

No I didn't, and I did notice it caused a hum so I just put a mic in there and that's stopped it (it was a tube mic power supply that was off). I tried with a resitor now and the diagrams were the same.

[quote author="NewYorkDave"]Is your heater supply floating--as in, it has no connection to ground or to any DC reference in the circuit? That'll definitely cause hum. Connect a 100-ohm, 1/2-watt resistor (preferably 1%) between each leg of the heater supply and ground. Assuming the layout was done well, there should not be excessive hum even with an AC heater supply. Did you twist the heater supply leads and route them away from signal leads?[/quote]

Yes the 100ohm resistors are there, just like in your PSU schematic (connected to chassis/ground). The leads are twisted and as far away from signal leads as the layout allowed. They cross several times, at 90 degree angles and do not touch at any point. I'd say the layout is pretty OK, but I'll see if I can optimise it slightly.

[quote author="NewYorkDave"]The 5965 was developed for computer use but has published characteristics similar to type 12AV7. And although I haven't used it myself, I know that many others have applied the 5965 successfully in audio circuits. Since your problem follows the tube, it stands to reason that there's probably something wrong with that tube--have you tried a different 5965? Is yours used or new-old-stock?[/quote]

It's NOS, and I didn't get a chance to try another one, yet. I'm getting a new one in a few days. Hope it turns out better.

Thanks for your time.

 

[khstudio]

Tubes can be funny so hopfully it's just that.

 

[NewYorkDave]

Sorry for the sloppy sketch... No time today to draft a proper diagram.

Recommended grounding scheme

Make certain that you ground pins 2 and 6 on the 5879, and bypass its cathode resistor with at least 47uF.

(I posted this also in Kiira's thread about her Two-Bottle build).

 

[Kingston]

Thanks newyorkdave. The grounding has always been as in your diagram.

It could be a problem with my layout. Maybe the "ball of caps" soldered in tube sockets isn't a good idea afterall. Maybe the transformers are picking up something they aren't supposed to. Could be the heater VAC crosses signal paths in a wrong way somewhere near the tube sockets.

Here it is.

Here's a high resolution version. http://www.michaelkingston.fi/files/twobottlebig.jpg

 

[NewYorkDave]

:shock:

Um... well... I'd really consider re-doing the layout.

 

[Kingston]

Hehe. yeah I thought it was pretty unorthodox and maybe bad, too. The signal paths are very short near the tubes though.

There's few things to test still, untill I resort to redoing it..

[EDIT]

I removed some unnecessary wires there that were unoptimal. some of the hum went away. S/N is around 85dB at minimun gain, although spectrum shows it could be -105dB, if no hum was present.

I got a new 5965A to test. It performs identically to the old one. A more experienced tube guy friend of mine will go through the unit with these and if we come up with a way to fix their performance I'll report here of course.

The two 12AV7 perform perfectly of course, and this unit is already in use. There are no serious problems on this anymore.

 

[khstudio]

Need some more help with 1 bottle Pre:

What is the BEST way to adapt (hook-up) this output transformer -

MODEL- MI-69
PRIMARY IMP- 150/600
SECONDAR-IMP- 15k
USE*- LI, LO
PRI POWERLEVEL 600- +18
MAGNETICSHIELDING- ?10 dB
CASES- M

I read this topic again and found info about what may have to be changed? R10 (220k) - Remove:?:
R8 330 - change to 1k :?:
Increace feedback resistor:?: Do I need to & where is it... the i bottle uses a pot for feedback, no? maybe the R12=10k?

Also, I'm just going to throw this together with the 5965's I have to get things going & I'll order 12av's BUT I'm worried -

Will 300+ volts be OK on these tubes:?: My PT is 250-0-250 & will end up a little higher than your original design.

 

[skipwave]

[quote author="khstudio"]Need some more help with 1 bottle Pre:

What is the BEST way to adapt (hook-up) this output transformer -

MODEL- MI-69[/quote]

I used MI-70 for the inputs on my One-Bottles. I have the pinout for the 69 at home. I'll check it and post the leads you should use later.

[quote author="khstudio"]Will 300+ volts be OK on these tubes:?: My PT is 250-0-250 & will end up a little higher than your original design.[/quote]

Are you powering two channels or one? Go back a few pages in this thread and you will find my questions and Dave's answers about modifying the power supply. It's all just Ohm's law! :green:

 

[khstudio]

[quote author="skipwave"][quote author="khstudio"]Need some more help with 1 bottle Pre:

What is the BEST way to adapt (hook-up) this output transformer -

MODEL- MI-69[/quote]

I used MI-70 for the inputs on my One-Bottles. I have the pinout for the 69 at home. I'll check it and post the leads you should use later.

[quote author="khstudio"]Will 300+ volts be OK on these tubes:?: My PT is 250-0-250 & will end up a little higher than your original design.[/quote]

Are you powering two channels or one? Go back a few pages in this thread and you will find my questions and Dave's answers about modifying the power supply. It's all just Ohm's law! :green:[/quote]

NO - I know what the leads are. I need help with the circuit changes that would go along with this OT transformer. I specified them above but I'm unsure, so I asked :wink:

POWER - My concerns are the tubes ability to handle the voltages, not the design.

 

[khstudio]

Skipwave, How did yours turn out?
Did you deviate from the original design or use an output trans?

 

[skipwave]

Yes, I did use output transformers, Triad A-69J. I really like the sound of the circuit driving the output transformers, a bit of extra color although too much on some sources which is why I hooked them up through a TRS switching jack allowing me to run unbalanced out. You don't need to make any changes to the circuit to use an output transformer.

The MI-69 is a microphone input transformer, but it might work ok in reverse. You would attach Green/White to 2.2uF output cap, Green and Brown to circuit ground, Yellow to XLR pin 2, Blue and Orange tied together, Black to XLR pin 3.

 

[NewYorkDave]

A 250-0-250V power transformer is perfect. Just use a fullwave rectifier (two diodes, transformer centertap grounded) instead of a fullwave bridge. The voltage will work out to be the same as if you had used a 250V secondary with a bridge. One nice thing about fullwave is that the transformer secondary only needs to be rated for 1.2 times the required DC output as opposed to 1.8 times with a fullwave bridge. A Google search will tell you all about the differences between fullwave and fullwave bridge. Here's one page that looks pretty straightforward.

 

[khstudio]

OK cool on the Power Supply... I got tied up in sessions but I'm back :green:

I'm confussed though, I read hear that I should do the folowing when using an output trans:

R10 (220k) - Remove
R8 (330) - change to 1k
Increace feedback resistor

But the problem is I'd like to use a switch to bypass the OT sometimes.
Should I just make a choice & go or will one setup optimally work for both?

The more I think about it, my LA-2A has a 10k input, so do my Disstressors & Fatso... & if I'm right, so does my Apogee & RME inputs -
so I guess the original output wouldn't be bad except for being unbal.

Still like to have the option though :grin:


Kevin

 

[NewYorkDave]

You misunderstood. The changes I was talking about (in another thread) had to do with using a gapped output transformer in series feed with no coupling cap. That requires a partial redesign of the circuit. But an ac-coupled output transformer doesn't; you're essentially just sticking the output transformer after the "stock" circuit.

 

[Kingston]

I took you advice early on and didn't put the 220K tiedown resistor in the preamp, as I permanently installed the OEP A262A2E as 4:1. http://www.oep.co.uk/audio_transformers_high_performance/audio_transformers_high_performance.html

It already sounds great, but I couldn't find out wether the transformer is gapped or not. I didn't switch the 330ohm plate resistor to 1k, and now I'm wondering if I should.

 

[NewYorkDave]

I'll try to make this more clear... If the output stage has a plate resistor and the output is coupled by a capacitor, don't change anything--whether your load is a gapped transformer, ungapped transformer or something else. The circuit only has to be changed if you're using a transformer primary in place of the plate resistor.